- 9,054
- Murica
- BasedAckbar
As much as 500 sounds nice, I won't hold my breath for that many to be added. If anything, it possibly means that GT Sport is not this smaller Gran Turismo game as originally thought.
Gran Turismo Sport: General Discussion
- Thread starter Formidable
- 47,132 comments
- 4,756,963 views
- 328
- United States
- pvvrry
I always guessed it was licensing for the lack of cars. That's how I kept my hopes up for more cars as DLC.
Last edited:
- 8,715
So the problem is a doubt about whether we could see 400 - 500 cars in GTS in a reasonable time frame?
It's not a question that can be answered. It's perfectly possible, of course.
As for the discussion itself, I see an issue with the use of the word "efficiency" - it means the ratio of useful product out to the effort (e.g. wage cost, time) put in. Employing exceedingly skilled artists to pore over photographs, CAD data and physical bits of the cars at length does not sound cheap.
The problem clearly hinges on the definition of "useful".
As for PD's modeling habits, I think it's clear that when recent games have launched they've had plenty of partially finished cars almost ready but not quite making the cut. I believe it was even mentioned in interview on occasion. That doesn't mean they do all planned cars at once. Cars are added to the queue (and allocated to a modeler) perhaps on an ad hoc basis according to the (re)licensing, then they pop out "when they're done".
Parallel work is indeed more "efficient" in general, but you have to pay attention to the detail: task switching.
Factories employ people to do the same task over and over again, because they build up a muscle memory and rhythm that allows sheer speed (efficiency = throughput per worker). Getting them all to switch roles every few minutes would kill the speed dead. That said, to prevent various types of fatigue (and injury...) roles are sometimes rotated on some longer timescale, with some adjustment period required (lower efficiency) that is worth the saving in whatever value the factory owner places on its workers' physical and mental wellbeing.
The artisan in their studio, however, will have many projects on the go and will tackle each sub-task associated with the projects according to when his or her mindset best matches what is required. This results in a much better quality product, but not finished very quickly, if at all in many cases (efficiency = how "good" the end product is compared to that of other makers).
It is likely in reality that PD sit somewhere closer to the factory worker paradigm than the true artisan, despite the shokunin aspect. But still, creative people work better when they can stay engaged, and task switching facilitates that. So that might mean they can model the headlights on one car, then move onto those for another they're working on until they get burned out on it, before switching to, say, wheels instead. That seems to me to be a reasonable compromise for a better end product, even if some of their effort doesn't make the initial cut (in the form of partially completed models). I think it's unlikely that anyone can sit at a single car and flesh it out from start to finish at optimum pace and quality; so, naturally, if forced to I'd wager it'd take them longer and / or look worse than if they were allowed to do what felt best at any given time.
Whilst any individual car may take a variable amount of time to make this way (from licensing to completion), effectively depending on how the artist "feels", you do eliminate "inefficiency" from certain bottlenecks. For example, not being able to flesh out the headlights' interior structure until they are finalised in terms of their shape and position in the car, which might be the case if you have people dedicated to certain smaller modeling tasks. That shouldn't happen with the correct ratio of "types" of modelers, but it would be extremely hard to balance over time, especially with a growing team.
The approach certainly doesn't stop certain cars being a priority, but it might also explain the Evora, and a hundred other examples of "why X and not Y?" from the past.
It's not a question that can be answered. It's perfectly possible, of course.
As for the discussion itself, I see an issue with the use of the word "efficiency" - it means the ratio of useful product out to the effort (e.g. wage cost, time) put in. Employing exceedingly skilled artists to pore over photographs, CAD data and physical bits of the cars at length does not sound cheap.
The problem clearly hinges on the definition of "useful".
As for PD's modeling habits, I think it's clear that when recent games have launched they've had plenty of partially finished cars almost ready but not quite making the cut. I believe it was even mentioned in interview on occasion. That doesn't mean they do all planned cars at once. Cars are added to the queue (and allocated to a modeler) perhaps on an ad hoc basis according to the (re)licensing, then they pop out "when they're done".
Parallel work is indeed more "efficient" in general, but you have to pay attention to the detail: task switching.
Factories employ people to do the same task over and over again, because they build up a muscle memory and rhythm that allows sheer speed (efficiency = throughput per worker). Getting them all to switch roles every few minutes would kill the speed dead. That said, to prevent various types of fatigue (and injury...) roles are sometimes rotated on some longer timescale, with some adjustment period required (lower efficiency) that is worth the saving in whatever value the factory owner places on its workers' physical and mental wellbeing.
The artisan in their studio, however, will have many projects on the go and will tackle each sub-task associated with the projects according to when his or her mindset best matches what is required. This results in a much better quality product, but not finished very quickly, if at all in many cases (efficiency = how "good" the end product is compared to that of other makers).
It is likely in reality that PD sit somewhere closer to the factory worker paradigm than the true artisan, despite the shokunin aspect. But still, creative people work better when they can stay engaged, and task switching facilitates that. So that might mean they can model the headlights on one car, then move onto those for another they're working on until they get burned out on it, before switching to, say, wheels instead. That seems to me to be a reasonable compromise for a better end product, even if some of their effort doesn't make the initial cut (in the form of partially completed models). I think it's unlikely that anyone can sit at a single car and flesh it out from start to finish at optimum pace and quality; so, naturally, if forced to I'd wager it'd take them longer and / or look worse than if they were allowed to do what felt best at any given time.
Whilst any individual car may take a variable amount of time to make this way (from licensing to completion), effectively depending on how the artist "feels", you do eliminate "inefficiency" from certain bottlenecks. For example, not being able to flesh out the headlights' interior structure until they are finalised in terms of their shape and position in the car, which might be the case if you have people dedicated to certain smaller modeling tasks. That shouldn't happen with the correct ratio of "types" of modelers, but it would be extremely hard to balance over time, especially with a growing team.
The approach certainly doesn't stop certain cars being a priority, but it might also explain the Evora, and a hundred other examples of "why X and not Y?" from the past.
- 3,264
- Salt Lake City,
Even if we only get half of that, say 250 total, that's still 110 more cars than what the game releases with. The "within a year" seems like it was a translation error. So the 400-500 "in total" sounds a lot more plausible. All this talk about taking it with a grain of salt and previous broken promises or delays in the past doesn't mean there's no dlc going to be added. If it's not the 400 amount at the end of the planned dlc, but they do add something like double the car total from launch, I think that's great. Even if we only reach 200 cars in all, still great imo. As long as they're not going to cost outrageous amounts, of course! 
- 11,660
- GTP_Orido
I have a feeling, PD have been making super premiums since at least 6 months or a year before PS4 release, those new GT6 premiums like Cizeta and Countach may have been downscaled super premium, they are exceptionally build compared to GT5 premiums. How long has it been since PS4 release ?
- 4,076
- Philippines
- GBalao888
- GBalao888
The PS4 was released two and a half years ago (three on November 15th) and is the same for GT6 (on December 6th). So it's been three years since the Cizeta V16T and the Countach have all been modeled to a high level of premium quality.
- 27,387
- Toronto
- NewAesthetic
- SlipZtrEm
Handily enough, GT Sport launches on the 3rd birthday of the PS4.
(EDIT) Ooof, soundly beaten by @GBalao888. I should learn to refresh tabs when they're left idle!
- 536
- United States
Even if we only get half of that, say 250 total, that's still 110 more cars than what the game releases with. The "within a year" seems like it was a translation error. So the 400-500 "in total" sounds a lot more plausible. All this talk about taking it with a grain of salt and previous broken promises or delays in the past doesn't mean there's no dlc going to be added. If it's not the 400 amount at the end of the planned dlc, but they do add something like double the car total from launch, I think that's great. Even if we only reach 200 cars in all, still great imo. As long as they're not going to cost outrageous amounts, of course!
Perhaps I have missed something, but has it been confirmed somewhere that the DLC requires a user to actually purchase it?
- 8,187
- Kaikoura, NZ
- Nato777
The time required per model depends I guess on what and how much has to be changed on the GT6 premiums to make them super premium. I don't know enough about 3D modeling or rendering to say but I suspect the Adaptive Tessellation and Physical Based Rendering are the main factors. In GT6 the AT was their own software based solution and I think more a 'pseudo' AT really. PS4 has hardware based AT so I would assume PD would have moved to use that instead, I have no idea if that affects the 3D model at all but at least I do know that not all premiums were capable of the AT in GT6 so those at least would need some work. The PBR to my understanding has mainly to do with the lighting engine.
Physical Based Rendering allows for much more accurate light simulation (reflection/diffusion/fresnel etc... off objects) with a more simplified instruction input to the engine. The data required by a PBR engine will be quite different to that of the GT6 engine for example meaning that if you take a GT6 car and plonk it into the GT Sport engine the rendering would be completely wrong or probably crash the system.
The GT6 models would need to be remodeled with the correct material/surface property data for the PBR engine to recognize. I have no clue what exactly that entails, if the material instruction set is embedded with each 'part' in the 3D model a complete rebuild would be in order. If the instruction set is via a global material map in the model data then it may only be that which needs updating.
The actual geometric data in the GT6 models may actually be perfectly fine - they were apparently built to a fidelity beyond that which the PS3 could display - just depends as I said on whether the AT requires a rebuild.
In short it's possible that reworking the GT6 premiums may be a relatively quick task, if so that begs the question: why aren't there many more of them in the launch product?? Maybe PD want to initially focus on the eSport mode of GT Sport and having a large car roster in the beginning would distract from that. Who knows, but it's kinda of fun/frustrating to speculate.
Physical Based Rendering allows for much more accurate light simulation (reflection/diffusion/fresnel etc... off objects) with a more simplified instruction input to the engine. The data required by a PBR engine will be quite different to that of the GT6 engine for example meaning that if you take a GT6 car and plonk it into the GT Sport engine the rendering would be completely wrong or probably crash the system.
The GT6 models would need to be remodeled with the correct material/surface property data for the PBR engine to recognize. I have no clue what exactly that entails, if the material instruction set is embedded with each 'part' in the 3D model a complete rebuild would be in order. If the instruction set is via a global material map in the model data then it may only be that which needs updating.
The actual geometric data in the GT6 models may actually be perfectly fine - they were apparently built to a fidelity beyond that which the PS3 could display - just depends as I said on whether the AT requires a rebuild.
In short it's possible that reworking the GT6 premiums may be a relatively quick task, if so that begs the question: why aren't there many more of them in the launch product?? Maybe PD want to initially focus on the eSport mode of GT Sport and having a large car roster in the beginning would distract from that. Who knows, but it's kinda of fun/frustrating to speculate.
- 27,387
- Toronto
- NewAesthetic
- SlipZtrEm
It hasn't. But it also hasn't been confirmed to be free either. I can't think of many triple-digit-car-roster car games that have at least doubled their count for free, though!
It's part of why Kaz' statement has stirred up so much discussion; there's quite a lot of variables at play here.
- 3,279
- France
http://www.autoevolution.com/news/s...ron-and-bugatti-vision-gt-concept-109985.html
Pretty crazy IMO. I imagine it will give some ideas to other manufacturers. Expect more VGTs
Pretty crazy IMO. I imagine it will give some ideas to other manufacturers. Expect more VGTs
- 1,492
- Germany
- KiroKai
- KiroKai
Is the difference really that big for Kaz?
Well, there's this:
If this applies to car modelling as well, what @Griffith500 assumes about freedom in the work process could be true. But I think Ridox might be right as well, if PD's internal communication is as great as claimed, shokunin might all be image and Kaz talk and in truth the single person approach to modelling a car is a thing of the past.
- 1,492
- Germany
- KiroKai
- KiroKai
It's gonna be exciting to see where he takes them and if we'll still see the Bugatti at future GT events/launch party.
mister dog
(Banned)
- 8,396
- Spain
- misterdog
- misterdog7
I wonder what the cars will be called for GT8; "Super duper premiums?".
If you don't need to distinguish anymore because you mix current and last gen models, then you could just as well call them 'cars' no?
If you don't need to distinguish anymore because you mix current and last gen models, then you could just as well call them 'cars' no?
- 27,387
- Toronto
- NewAesthetic
- SlipZtrEm
I'd suggest PD could adopt the Street Fighter approach, but I don't know how well that'd go over with fans. We've got Super Premiums, and we could definitely get Ultra Premiums. But people might balk at Super Premium: Arcade Edition.
mister dog
(Banned)
- 8,396
- Spain
- misterdog
- misterdog7
- 11,660
- GTP_Orido
By GT10, the models would be called "Masterpiece Premium", unrivaled in accuracy and details 
I think GTS super premium would be the best 3d car models in console when in highest LOD ( showroom view )
One of my concern is clutch button implementation, PS2 drift games have it, why would GTS not have it ? I want to use clutch when using DS4, clutch kick yo !! It may be useful in certain situation ( not drifting only ). Doing driving stunt moves is one of them ( with clutch ), a great test of a good sim that I will try in AC as well. Gymkhana is the next feature I have been waiting since TXRD PS2, car behavior deficiency easily seen when doing those extreme maneuver.
I think GTS super premium would be the best 3d car models in console when in highest LOD ( showroom view )
One of my concern is clutch button implementation, PS2 drift games have it, why would GTS not have it ? I want to use clutch when using DS4, clutch kick yo !!
It may be useful in certain situation ( not drifting only ). Doing driving stunt moves is one of them ( with clutch ), a great test of a good sim that I will try in AC as well. Gymkhana is the next feature I have been waiting since TXRD PS2, car behavior deficiency easily seen when doing those extreme maneuver.
- 396
- Houston, TX
- realtexan
I doubt super premiums are an official term for the cars in GT Sport. It's just used to describe the quality of the cars in GT Sport compared to GT5 and GT6.
- 536
- United States
I think this is not only exciting, but it validates the VGT Project even further. Not only are there real world manufacturers involved and prototypes being produced, now it seems there is some serious money involved it.
I fully expect other racing games out there to jump in this unique approach that unites the auto industry to the video game world.
mister dog
(Banned)
- 8,396
- Spain
- misterdog
- misterdog7
- 4,089
- Utah
I was kind of excited to hear about the possible 400-500 cars and I though "Gran Turismo is finally back!", but honestly, I won't believe it until I see it. If they really put their mind to it and do it within a reasonable time line, I wouldn't even mind paying for extra content no matter how much it is. As long as PD decides to revive GT and stop experimenting.
For now I've got the holy trinity of car games: Forza 6 for eye and ear candy, Forza Horizon 3 for that open-world driving experience and Assetto Corsa for a real physics thrill.
For now I've got the holy trinity of car games: Forza 6 for eye and ear candy, Forza Horizon 3 for that open-world driving experience and Assetto Corsa for a real physics thrill.
- 8,715
If PD were smart, they'd have tailored the content to work with hardware tessellation units as well, since their software solution wouldn't be limited algorithmically, just in terms of how to parcel it up on the SPEs.The time required per model depends I guess on what and how much has to be changed on the GT6 premiums to make them super premium. I don't know enough about 3D modeling or rendering to say but I suspect the Adaptive Tessellation and Physical Based Rendering are the main factors. In GT6 the AT was their own software based solution and I think more a 'pseudo' AT really. PS4 has hardware based AT so I would assume PD would have moved to use that instead, I have no idea if that affects the 3D model at all but at least I do know that not all premiums were capable of the AT in GT6 so those at least would need some work. The PBR to my understanding has mainly to do with the lighting engine.
Physical Based Rendering allows for much more accurate light simulation (reflection/diffusion/fresnel etc... off objects) with a more simplified instruction input to the engine. The data required by a PBR engine will be quite different to that of the GT6 engine for example meaning that if you take a GT6 car and plonk it into the GT Sport engine the rendering would be completely wrong or probably crash the system.
The GT6 models would need to be remodeled with the correct material/surface property data for the PBR engine to recognize. I have no clue what exactly that entails, if the material instruction set is embedded with each 'part' in the 3D model a complete rebuild would be in order. If the instruction set is via a global material map in the model data then it may only be that which needs updating.
The actual geometric data in the GT6 models may actually be perfectly fine - they were apparently built to a fidelity beyond that which the PS3 could display - just depends as I said on whether the AT requires a rebuild.
In short it's possible that reworking the GT6 premiums may be a relatively quick task, if so that begs the question: why aren't there many more of them in the launch product?? Maybe PD want to initially focus on the eSport mode of GT Sport and having a large car roster in the beginning would distract from that. Who knows, but it's kinda of fun/frustrating to speculate.
The "adaptive" part is two way, and the hardware tessellators only work one way: subdivision. The other part is the progressive mesh system which will work on any platform you can code stream processors for. So the raw geometry assets likely don't need to be any different from the "super premiums" and the tessellated premiums on PS3.
The "physically based renderer" stuff (assuming any of it is technically new) needs work in terms of the new shaders they might be running on PS4 vs. those on PS3, but the cars were already modeled in terms of materials and shaders anyway with little reliance on textures except for detail work (and decals), and even then they clearly made use of separate channels for the various "physical properties". So that's not quite scratch work, either.
So the work is mainly in making the cars tessellation-ready (existing Premiums could be done quicker with custom tools), as well as preparing them for the livery editor.
- 27,387
- Toronto
- NewAesthetic
- SlipZtrEm
It certainly proves VGT is a valid marketing approach. But then again, I think most people would've already agreed it was essentially advertising anyway.
Though a rich Arabian prince being able to wave enough money at a company to get what he wants is hardly new!
- 137
- Japan
Kaz has said, "200 people of the staff" in E3.
Data of the ending movie and PD official (was a total of 110 people) it is old.
However, I do not think that about 90 people were all in the modeling team. Many are in about 30 people, then are distributed to each department. So, it would not is greatly increased.
